Analytics Friday

The episode contains a handful of genuinely useful ideas - decision laundering, the negotiable vs. non-negotiable data analogy with finance, and agent-tracking as a future problem - but is heavily padded by host rambling, mutual agreement loops, and unclear questions that diffuse rather than deepen the ideas.

“analytics data is that somehow we ended up in a box where analytics data is negotiable”

“your data is always wrong. The important thing that is that it's equally wrong all the time because then the trends are correct”

'Decision laundering' is a genuinely original and useful frame for a common dysfunction, and the finance-vs-analytics accountability analogy is a crisp reframing; but much of the rest - proactive analysts, culture over tools, data quality problems - is well-worn territory in the analytics practitioner community, and the Bezos 70% anecdote and Steve Jobs pirate quote are recycled.

“people in organizations are getting really good at using analytics to justify decisions they have made without data”

“you will not go to the finance team and ask for better numbers”

Steen Rasmussen has genuine long-term practitioner credentials - co-founder of a leading Nordic analytics agency for 20 years, multiple industry nominations - and speaks from real operator experience rather than thought-leadership abstraction, though he is a regional expert rather than a globally recognised name.

“IH Nordic1, um, best analytics agency in Denmark, uh, 12 years in a row. We got nominated five times by the Digital analytics association as the best analytics agency in the world”

“I started co founded an agency 20 years ago called IH Nordic”

The episode names real tools and frameworks (GA4 vs Universal Analytics, MMM/Meridian, UTM codes, Measure Camp) and cites the Bezos 70% heuristic, but there are no client case studies, no revenue or ROI figures, no named campaigns, and no concrete before/after data to ground the claims.

“the biggest problem they have uh, in general is that they can. I'll say it, I'll. They suck at campaign tracking. They're really bad at using UTM codes”

“the cost of the next percentage of perfection, uh, is increasing. So saying going from 80 to 81% that cost uh, something. But going from 81 to 82 that costs more”

The host's questions are frequently multi-part, self-answering, and grammatically unclear, making it hard for the guest to give focused answers; there is almost no pushback or productive disagreement, and the conversation frequently collapses into mutual validation rather than deeper interrogation of the guest's claims.

“So um, um, you see amplitude and you see other tools because of the gdpr, uh, rising and uh, it's okay with that. But um, what you see about um, the customers, they are seeing the value of that or um, they are constantly silos”

“Yeah, uh, that's, that's the way”